Eltec 406MCA Tester

Code for the 406MCA sensitivity, polarity, offset, noise, and SNR test process. The tester apps use the LabJack T7-Pro through the labjack.ljm Python package and include a simulator mode, so the operator flow and analysis can be exercised without hardware connected.

The repo holds four versions of the same technician data-collection app (one per physical test setup, plus the v4 visual refresh of the emitter rig app), the engineer tools used to bring up and validate new setups, and per-version analysis scripts.

Repository layout

tech_app/                         Technician data-collection app, by version
├── v1_single_sensor/             Original single-sensor rig with the AM502 amplifier
│   ├── eltec_406mca_tester.py        (also the shared signal-math + LabJack engine)
│   └── Run 406MCA Tester.bat
├── v2_scope_verification/        Older scope-verification rig (guided lot flow)
│   ├── eltec_406mca_scope_verification_tester.py
│   └── Run 406MCA Scope Verification Tester.bat
├── v3_emitter/                   LabJack-driven emitter rig, no AM502 (unity-gain buffer)
│   ├── eltec_406mca_emitter_tester.py
│   ├── Run 406MCA Emitter Tester.bat
│   ├── Create Desktop Shortcut.ps1
│   └── assets/README.txt
└── v4_emitter/   ← CURRENT       Same emitter rig + measurement engine as v3, with the
    ├── eltec_406mca_emitter_tester.py   Eltec-branded animated UI (eltecinstruments.com look)
    ├── Run 406MCA Emitter Tester.bat
    ├── Create Desktop Shortcut.ps1
    └── assets/README.txt         (optional brand fonts go in assets\fonts\)

engineer_tools/                   Engineer setup / bring-up tools (live signal monitor)
├── eltec_406mca_signal_monitor.py
├── eltec_406mca_signal_monitor_ui.py
└── Run 406MCA Signal Monitor UI.bat

analysis/                         Analysis, by app version
├── v1_single_sensor/            analyze_406mca_snr_results.py, test_406mca_analysis.py
├── v2_scope_verification/       dataAnalysis.py, disagreementAnalysis.py
├── v3_emitter/                  analyze_emitter_results.py + Run Emitter Analysis.bat
├── v4_emitter/                  analyze_emitter_results.py + Run Emitter Analysis.bat (v4 data)
└── reports/                     Reference / generated documents (buffer + SNR write-ups)

assets/eltec_logo.png            Shared logo used by the apps

eltec_406mca_tester.py (v1) doubles as the single source of truth for the signal math and the LabJack device wrapper; the v3/v4 emitter testers and both engineer tools import from it, so it is kept alongside the other versions rather than deprecated.

Where test data is saved

Each version writes into its own subfolder of the results root so data can be tracked and analyzed per version:

C:\Users\<user>\Documents\Eltec_406MCA_Test_Results\
├── v1_single_sensor\   406mca_results.csv
├── v2_scope_verification\
│   ├── 406mca_scope_verification_lot_<lot>.csv
│   ├── autosave\  waveform_snapshots\  analysis\
├── v3_emitter\
│   ├── 406mca_emitter_lot_<batch>.csv
│   ├── autosave\  waveform_snapshots\  analysis\
└── v4_emitter\
    ├── 406mca_emitter_lot_<batch>.csv
    ├── autosave\  waveform_snapshots\  analysis\

Each app creates its subfolder automatically on first save.

---

v4 – Emitter Tester (current, visual refresh)

tech_app/v4_emitter/eltec_406mca_emitter_tester.py is the same rig, wiring, guided flow, measurement engine, and CSV schema as v3 with a complete visual overhaul styled after eltecinstruments.com:

• Eltec-blue gradient app bar with the logo badge, an animated signal-trace line, and a live battery gauge pill (click it to re-check the 9V supply).
• Numbered step rail (01 / 02 / 03) with animated check-offs and connectors.
• Rounded soft-shadow cards with the Eltec technical-gradient accent strip, hover-animated rounded buttons, and an animated waveform toggle switch.
• Animated PASS/FAIL banner, count-up result tiles, and a scanning progress bar while a measurement runs.
• Dark navy oscilloscope panel (grid + glow traces + sweep beam) for the live AIN0/AIN2 view, with monospace technical readouts.

Optional: drop Poppins/Manrope/JetBrains Mono .ttf files into tech_app\v4_emitter\assets\fonts\ and the app loads them privately at startup for an even closer match to the website type; otherwise it falls back to Segoe UI / Consolas automatically (see assets\README.txt).

Capture modes (v4 speed-up)

v3 needs 45-80 chopping cycles (4.5-8 s) per sensor because its stability rule compares two 20-cycle averages. v4 adds a margin-based early exit plus shorter DC reads (offset 24x3 ms instead of 80x10 ms; the battery check is reused when under 30 s old). The Capture mode selector under Advanced options picks:

• Fast (early exit) — stops as soon as the waveform is stable over two 6-cycle windows AND every metric is decisively clear of its limit (sensitivity ≥1.5x/≤0.5x the minimum, polarity confidence ≥1.5x threshold, SNR ≥2x/≤0.5x the gate), confirmed on two consecutive cycles. Clear sensors decide in ~1.5-2 s; marginal ones keep capturing up to a 30-cycle cap.
• Validation (default) — runs the full v3-length capture AND logs what Fast would have decided to the fast_* CSV columns plus a fast_match YES/NO. Run production batches in this mode first; the batch summary shows a "FAST PATH n/m MATCH" chip. Switch to Fast only after mismatch-free runs.
• Full (v3 timing) — exact v3 behavior.

The FAST_* constants at the top of the tester set the windows and decision margins; tune them from validation-mode data. New CSV columns: capture_mode, capture_cycles, capture_seconds, fast_stop_cycle, fast_sensitivity_mv, fast_polarity, fast_pass_fail, fast_match, data_source (older batch CSVs keep their original columns automatically).

"Is everything plugged in?" guards (v4)

v3 silently switched to simulator mode when no T7 was found, so a technician could unknowingly record synthetic numbers (battery pinned at 8.8 V). v4:

• Simulator is explicit opt-in (Advanced options), shows an amber SIMULATOR badge in the header, marks the result detail line "SIMULATED DATA", and tags CSV rows data_source=simulator.
• No T7 detected — battery shows -- and Measure explains to plug in the LabJack instead of running.
• Battery/AIN1 wiring fault — readings outside 3.0-10.5 V (floating input, missing battery clip) show a red "CHECK WIRING" pill, a red banner, and block the Measure button.
• Sensor pre-flight — before capturing, the DC offset must be in the 0.05-2.5 V plausible band; otherwise the test aborts with "No sensor detected - seat the sensor in the rig" and nothing is recorded.
• PWM sync pre-flight — a ~0.3 s peek at AIN2 after the PWM starts must see a square wave; otherwise the test aborts naming the DIO0/AIN2 wiring.

Results are logged to …\v4_emitter\406mca_emitter_lot_<batch>.csv (v3 columns + the capture telemetry above). Run it with Run 406MCA Emitter Tester.bat or:

cd C:\Users\vma\Documents\Eltec406MCATester\tech_app\v4_emitter
python eltec_406mca_emitter_tester.py

Create Desktop Shortcut.ps1 works the same as v3's and names the shortcut "Eltec 406MCA Emitter Tester v4".

---

v3 – Emitter Tester (previous, technician-friendly)

tech_app/v3_emitter/eltec_406mca_emitter_tester.py is the step-by-step app for the rig where the LabJack drives the emitter itself and the sensor is read without the AM502 amplifier. A unity-gain (voltage-follower) op-amp buffer feeds the LabJack a low-impedance signal while preserving the ~0.667 V DC offset and the small AC waveform, so the external gain is always 1x.

Wiring for this rig

• AIN0: buffered sensor signal — carries BOTH the DC offset and the AC signal.
• AIN2: PWM / MOSFET-gate drive, looped back as the polarity/sync reference.
• DIO0: PWM output to the MOSFET gate that switches the emitter. Use a common ground with the emitter supply.

The LabJack generates a 10 Hz PWM (default DIO0, 50% duty) to switch the MOSFET that drives the emitter. Because the PWM is also wired into AIN2, the rising-edge polarity check works unchanged. Use the +/-1 V (x10) AIN0 range so the ~0.667 V offset plus the small AC signal fit with good resolution.

Guided flow

1. Enter the batch number, tester name, and filter/setup, then press Enter (the batch field is focused on launch, so you can type immediately).
2. Place the sensor in the rig and press Enter.
3. The app reads the DC offset (emitter off), turns the PWM emitter on, measures sensitivity and polarity, and shows the offset, sensitivity, and a GOOD/BAD polarity verdict. The screen turns green for PASS or red for FAIL.
   • Comment records a note for the sensor.
   • Capture waveform saves a PNG of the AIN0/AIN2 traces under the version's waveform_snapshots\ folder.
   • Show waveform reveals the live AIN0 + AIN2 traces while it reads.
4. Save + Next Sensor (Enter) auto-increments the sensor number; Save + Exit Batch (Esc) saves and shows a batch summary.

Results are logged per batch to …\v3_emitter\406mca_emitter_lot_<batch>.csv with columns: timestamp, batch_number, sensor_number, sensor_id, tester_name, model, filter_setup, pwm_channel, pwm_hz, pwm_duty, offset_v, sensitivity_mv, polarity, polarity_good_bad, pass_fail, fail_reasons, operator_comments, waveform_snapshot_paths, battery_v, noise_rms_mv, snr_db.

If no T7 is detected the app drops into simulator mode so the full flow can be walked through without hardware.

Run it / make a desktop icon

cd C:\Users\vma\Documents\Eltec406MCATester\tech_app\v3_emitter
python eltec_406mca_emitter_tester.py

Or double-click Run 406MCA Emitter Tester.bat. To put a clickable ELTEC-logo icon on the desktop, run once:

cd C:\Users\vma\Documents\Eltec406MCATester\tech_app\v3_emitter
powershell -ExecutionPolicy Bypass -File ".\Create Desktop Shortcut.ps1"

It builds a multi-size eltec_logo.ico from assets\eltec_logo.png using built-in Windows imaging (no Pillow needed) and points the shortcut at it. Re-run it if the app folder moves or the logo changes.

The logo loads from tech_app\v3_emitter\assets\eltec_logo.png or the shared repo-root assets\eltec_logo.png; if neither is present a drawn ELTEC logo is used.

---

v2 – Scope Verification Tester

tech_app/v2_scope_verification/eltec_406mca_scope_verification_tester.py is the guided lot-based app used on the older scope-verification rig. It logs per lot to …\v2_scope_verification\406mca_scope_verification_lot_<lot>.csv.

cd C:\Users\vma\Documents\Eltec406MCATester\tech_app\v2_scope_verification
python eltec_406mca_scope_verification_tester.py

---

v1 – Single-Sensor Tester (original, AM502)

tech_app/v1_single_sensor/eltec_406mca_tester.py is the first single-sensor program, reading the sensor through the AM502 amplifier. It also provides the Distance cm / Input voltage V sweep fields used for the SNR study. It logs to …\v1_single_sensor\406mca_results.csv.

cd C:\Users\vma\Documents\Eltec406MCATester\tech_app\v1_single_sensor
python eltec_406mca_tester.py

Default wiring

• AIN0: sensor output or conditioned waveform signal.
• AIN2: blade sync signal.

The procedure uses the rising edge of AIN2 as the polarity reference. The tester does not offer falling-edge polarity testing in the operator UI. Close LabJack programs such as LJStreamM or Kipling before using hardware mode — only one process can claim the T7 USB connection at a time.

---

Engineer tools

engineer_tools/eltec_406mca_signal_monitor.py (CLI) and eltec_406mca_signal_monitor_ui.py (Tkinter) watch the incoming LabJack signal continuously instead of taking one pass/fail snapshot. They record a baseline and report whether the current AIN0 waveform has changed enough to confirm the setup is responding — useful when bringing up or debugging a new emitter/sensor rig. Both reuse the signal math from the v1 tester.

cd C:\Users\vma\Documents\Eltec406MCATester\engineer_tools
python eltec_406mca_signal_monitor_ui.py          # or: Run 406MCA Signal Monitor UI.bat
python eltec_406mca_signal_monitor.py --simulator  # CLI, no hardware

---

Analysis

Each analyzer defaults to its version's results subfolder and writes reports into an analysis\ folder there (CSV exports + a self-contained HTML report).

# v4 emitter (current): per-lot yield, offset/sensitivity/SNR stats, failure reasons, outliers
cd C:\Users\vma\Documents\Eltec406MCATester\analysis\v4_emitter
python analyze_emitter_results.py                 # or: Run Emitter Analysis.bat

# v3 emitter: same analysis over the v3 results folder
cd C:\Users\vma\Documents\Eltec406MCATester\analysis\v3_emitter
python analyze_emitter_results.py                 # or: Run Emitter Analysis.bat

# v2 scope verification: lot summaries, program-vs-operator disagreements, outliers
cd C:\Users\vma\Documents\Eltec406MCATester\analysis\v2_scope_verification
python dataAnalysis.py
python disagreementAnalysis.py

# v1 single-sensor: SNR distance/voltage sweep report (+ Word .docx) and math self-tests
cd C:\Users\vma\Documents\Eltec406MCATester\analysis\v1_single_sensor
python analyze_406mca_snr_results.py
python test_406mca_analysis.py

All analysis scripts use only the Python standard library so they run on the tester PC without extra packages. Pass --results-dir / --output-dir to override the defaults.

---

Current 406MCA limits

• Frequency setup: 10 Hz +/- 0.1 Hz.
• Offset: 0.3 Vdc minimum, 1.2 Vdc maximum.
• Polarity: positive for a positive change.
• Sensitivity minimum:
  • -3 filter: 25 mV
  • -27 filter: 25 mV
  • -266 filter: 30.9 mV
  • -273 filter + blackened tube: 2.3 mV
  • -284 filter + extra -6 + blackened tube: 4.0 mV

Measurement notes

The offset is estimated from the average voltage of the AIN0 waveform during the same stable, complete blade-sync cycles used for the sensitivity reading. The waveform is expected to be triangular. The program segments the signal by blade-sync cycles and watches the cycle peak-to-peak readings until the rolling average is stable within 10%, then uses the median peak-to-peak voltage from the stable cycles. The signal gain defaults to 1x for direct sensor readings; if an amplifier is used, enter its gain so the program divides the measured waveform before comparing against the sensitivity limit.

The LabJack AIN range setting is an input range, not an external gain correction. LJM returns calibrated volts, so leave external gain at 1x when reading the sensor directly even if AIN0 is set to the +/-1 V (x10) range. During stream, the app samples AIN0 multiple times after the sync channel and keeps the final reading to reduce settling error from high-impedance sensor outputs.

Polarity is measured against the rising blade-sync edge. The estimator searches the early rising-edge response region for the strongest signed waveform change and reports the response phase window and confidence as a percentage of the cycle peak-to-peak value.

Noise is estimated from the same stable blade-sync cycles used for sensitivity. The app aligns the stable cycles by phase, subtracts the average cycle shape, and reports the residual RMS as gain-corrected noise in millivolts. SNR is reported in dB from signal RMS divided by noise RMS.

If the measured AIN0 average is near 0 V, the waveform output is not carrying the detector offset. Check that the waveform output path is DC-coupled and referenced to the same ground. Frequency, clipping, and stability issues are shown as warnings; the pass/fail decision is based on sensitivity, polarity, offset, and whether the waveform stabilized before the capture limit.

Finding the best input voltage for SNR (v1 sweep study)

Use the same sensor, filter/setup, gain, LabJack range, wiring, emitter drive method, and fixture alignment for the whole comparison. For each distance, sweep the same voltage points, and enter each run's Distance cm and Input voltage V so the CSV can be grouped later.

1. Choose a safe voltage sweep before starting. Stay within the sensor, emitter, fixture, and LabJack input limits. If the safe range is unknown, start low and increase in small steps.
2. Pick 5 to 10 voltage points across the range. Smaller steps near the expected best voltage are useful.
3. Pick the distances to test (such as 45 cm, 55 cm, 65 cm) and use the same voltage list at every distance.
4. Set the first distance, enter it in Distance cm, then run the full voltage sweep.
5. At each voltage, wait for the fixture and waveform to settle before Start Test.
6. Run at least 3 tests per voltage/distance (5 if readings vary a lot).
7. Move to the next distance and repeat the same voltage sweep.
8. Watch for clipping, unstable-waveform, offset, or polarity warnings. Do not treat a voltage as best if it only looks good because the waveform is clipping or unstable.
9. After the first sweep, repeat the best few combinations in reverse order to catch drift from warm-up, sensor heating, or setup changes.
10. Compare each combination by average snr_db (higher is better); also check noise_rms_mv, sensitivity_mv, pass/fail, and warnings.
11. If two settings are within ~1–2 dB, prefer the lower, safer voltage and easier fixture distance unless production requires otherwise.

analysis/v1_single_sensor/analyze_406mca_snr_results.py groups these sweep rows and writes a console summary, a group-summary CSV, and a Word .docx report. Useful columns: distance_cm, input_voltage_v, sensor_id, filter_setup, sensitivity_mv, noise_rms_mv, snr_db, pass_fail, fail_reasons.

Safety notes

• Keep the AIN0 signal within the LabJack analog input range.
• Ensure the fixture, sensor/LabJack input, and sync signal share a valid common ground.
• If the T7 is claimed by another program, close LJStreamM/Kipling and press Connect again.